Log loss scenario of internet of vehicles connection abnormal probability compensation positioning method and system
By constructing a set of compensation features in the vehicle networking system and calculating anomaly scores using proximity time series and group statistical features, the problem of accurate location of connection anomalies caused by missing logs was solved, an automated fault handling process was realized, and the operation and maintenance efficiency of large-scale fleets was improved.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- SHANGHAI YOUKA NETWORK TECH CO LTD
- Filing Date
- 2026-05-11
- Publication Date
- 2026-07-03
Smart Images
Figure CN122339957A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to a method and system for compensating for abnormal connection probability in vehicle network scenarios with missing logs. Background Technology
[0002] With the large-scale deployment of vehicle-to-everything (V2X) terminals in passenger cars, commercial vehicles, and special-purpose vehicles, these vehicles typically perform data reporting, remote diagnostics, remote control, and online monitoring through TBOX (On-Board Module), communication modules, SIM or eSIM (Embedded Subscriber Identity Module), operator wireless networks, and cloud access platforms. The stability of the connection status and the efficiency of anomaly localization directly affect the continuity of V2X services, platform operation and maintenance costs, and vehicle service quality.
[0003] Current methods for locating connectivity anomalies in vehicle-to-everything (V2X) networks typically rely on complete logs reported by vehicle terminals, attachment information returned by modules, platform-side access logs, and manual experience for troubleshooting. When complete logs are available, maintenance personnel can determine the cause of the anomaly based on fault codes, network registration processes, heartbeat records, APN configurations, authentication results, or historical case libraries.
[0004] However, in real-world operations and maintenance environments, connection anomalies often occur in scenarios such as weak networks, network outages, terminal restarts, frequent network switching, module malfunctions, or storage overwrites, causing the cloud platform to be unable to collect complete logs at critical moments. In such cases, traditional methods that rely on complete logs for troubleshooting are insufficient to provide effective conclusions, forcing operations and maintenance personnel to manually investigate using fragmented information, which is time-consuming and results in inconsistent judgments.
[0005] In existing technologies, the analysis of abnormal vehicle-to-everything (V2X) connections typically follows these steps: First, after the platform detects that a vehicle has been offline for an extended period, its heartbeat has been interrupted, or its attachment has failed, it collects existing logs from both the terminal and platform sides. Second, the collected logs are matched with preset rules, fault trees, or historical case libraries. Third, based on the matching results, the possible fault types are determined, and troubleshooting suggestions or recovery instructions are generated.
[0006] In the aforementioned technical approaches, existing technologies generally assume that all key logs have already been collected, such as vehicle attachment process logs, SIM or eSIM status logs, network access status logs, and platform authentication logs. Even if some solutions incorporate historical cases or knowledge bases, they typically perform matching and analysis based on the existing key logs.
[0007] Existing technologies lack a dedicated processing mechanism for "missing or incomplete critical logs." In other words, when a terminal is unable to report complete logs, there is a lack of a technical solution to compensate for and infer the cause of the anomaly by using alternative information such as adjacent time-series status, recent successful connection characteristics, and statistical characteristics of similar vehicle groups in the same area.
[0008] Currently, the main drawbacks are as follows: (1) Over-reliance on complete logs. Most existing solutions require relatively complete terminal logs, registration logs, authentication logs or platform access logs as input. Once key fields are missing, the accuracy of location drops significantly.
[0009] (2) Lack of ability to compensate for missing information. For situations such as log loss, overwriting, and breakpoint reporting, existing technologies can often only mark them as "awaiting manual analysis" and cannot provide probabilistic location results based on alternative information.
[0010] (3) Insufficient support for the operation and maintenance of large fleets. Vehicle networking platforms often manage a large number of vehicles at the same time. If a large number of anomalies rely on manual verification, it will cause a backlog of work orders and a decrease in the efficiency of fault handling.
[0011] (4) The location results are difficult to directly connect with operation and maintenance actions. Existing technologies mostly remain at the level of "anomaly detection" or "rule hit", and have not formed a linkage mechanism between candidate cause ranking and subsequent handling process.
[0012] Therefore, to address the above issues, a method and system for compensating for abnormal vehicle network connections in scenarios with missing logs are provided. Summary of the Invention
[0013] To address the aforementioned problems in existing technologies, this invention provides a method and system for compensating for abnormal vehicle network connections in scenarios with missing logs. By adding a log integrity determination and compensation positioning mechanism, even when critical logs are missing or incomplete, it can still output candidate abnormality cause ranking results based on alternative information, thus reducing the reliance of existing technologies on complete logs.
[0014] The technical solution to achieve the above objectives is: One of the present inventions provides a method for compensating for the probability of abnormal vehicle network connections in scenarios with missing logs, comprising: Step S1: Obtain multi-source data related to connection anomalies of the target vehicle-to-everything (V2X) terminal within a preset time window; Step S2: Perform integrity checks on the multi-source data to determine whether there are missing logs related to connection anomalies, and determine the type and extent of missing logs; Step S3: When it is determined that there are missing logs, extract the adjacent temporal state features of the target vehicle network terminal within a preset time period before and after the anomaly occurred; Step S4: Extract the historical successful connection features of the target vehicle-to-everything (V2X) terminal; Step S5: Filter the reference terminal group that has preset similar conditions to the target vehicle-to-everything (V2X) terminal and extract the group's statistical characteristics; Step S6: Construct a set of compensation features based on the characteristics of adjacent time series states, historical successful connections, and group statistical features; Step S7: Based on the compensation feature set, calculate the anomaly score of each cause in the candidate anomaly cause set; Step S8: Normalize the anomaly score into a probability value to generate an anomaly localization result that includes candidate cause ranking, probability value, and confidence information; Step S9: Trigger the handling process according to the probability value sorting; the handling process includes automatic recovery, log supplementation, re-issuance of configuration, generation of work order and transfer to manual review.
[0015] Preferably, in step S1, the multi-source data includes at least one or more of the following: terminal status data, network connection data, platform access data, and historical connection records.
[0016] Preferably, in step S2, the missing log types include, but are not limited to, missing attachment logs, missing authentication logs, missing heartbeat logs, missing network switching logs, missing eSIM status logs, and missing platform access logs.
[0017] Preferably, in step S3, the adjacent time-series state features include, but are not limited to, the time of the last successful heartbeat before the anomaly, the most recent successful registration result, the number of reconnections before the anomaly, and the most recent network switching result.
[0018] Preferably, in step S4, the characteristics of historical successful connections include, but are not limited to, the operator identifier, access point name, IP allocation result, connection establishment duration, continuous online duration, network type, and signal strength distribution in the most recent successful session.
[0019] Preferably, in step S5, similar conditions include, but are not limited to, one or more of the following: same region, same operator, same vehicle model, same terminal version, same module version, and same time window.
[0020] Preferably, in step S7, the candidate abnormality reasons include at least one or more of the following: terminal-side abnormality, communication module abnormality, SIM or eSIM abnormality, network access abnormality, operator-side abnormality, and platform-side access abnormality. Let the set of candidate anomaly causes be . Based on the compensation feature set, the matching degree between each preset candidate anomaly cause and the compensation feature set is evaluated to obtain the candidate anomaly causes. Corresponding abnormal scores ; Among them, the compensation features are derived from the features of adjacent time-series states, the features of historical successful connections, and the statistical features of the reference terminal group; Abnormal scores satisfy: ; In the formula, For candidate anomaly causes, use adjacent time-series state features The level of support Historical successful connection features for candidate anomaly causes The level of support To identify candidate anomaly causes based on the statistical characteristics of the terminal group The level of support Preset weighting coefficients for adjacent temporal state features. Preset weighting coefficients for historically successful connection features. To reference the preset weighting coefficients of the statistical characteristics of the terminal group, , and All are greater than 0.
[0021] Preferably, in step S8, the anomaly score of each candidate anomaly cause is calculated. Normalization yields the probability value : .
[0022] Preferably, in step S9, when When the value exceeds a preset threshold, it can be directly triggered based on the candidate anomaly cause. The corresponding automatic recovery action; when When the value is below the preset threshold, the process of supplementing logs, re-issuing configurations, generating work orders, or transferring the process to manual review is triggered.
[0023] A second aspect of the present invention provides a vehicle network connection anomaly probability compensation and positioning system for implementing a vehicle network connection anomaly probability compensation and positioning method in a scenario of missing logs, comprising: The anomaly detection module is used to detect whether the target vehicle has connection anomalies. Anomalies include, but are not limited to, prolonged offline status, interrupted heartbeat, registration failure, authentication failure, frequent reconnection, or loss of connection after switching networks. The data acquisition module is used to acquire multi-source data related to connection anomalies within a preset time window when connection anomalies occur. The multi-source data includes at least one or more of the following: terminal status data, network connection data, platform access data, and historical connection records. The integrity determination module is used to identify the type and extent of missing logs and determine whether to proceed with the compensation and location process. The compensation feature construction module is used to construct a set of compensation features based on the adjacent time-series status, historical successful connection features, and statistical features of similar vehicle groups when there are missing logs. The probability localization module is used to calculate the probability value of multiple candidate anomalies, generate anomaly localization results that include the candidate cause ranking, probability value and confidence information, and sort them according to the probability value. The results output and processing linkage module is used to trigger recovery actions, supplement logs, generate work orders, or transfer to manual review based on the sorting results.
[0024] Compared with the prior art, the beneficial effects of the present invention are: 1) By adding a log integrity determination and compensation location mechanism, this invention can still output candidate anomaly cause ranking results based on alternative information when key logs are missing or incomplete, thus reducing the dependence of existing technologies on complete logs; 2) This invention utilizes adjacent time-series states, historical successful connection features, and statistical features of similar terminal groups to construct a set of compensation features. Compared with relying solely on a single log fragment, this method can improve the stability and accuracy of connection anomaly judgment in missing scenarios. 3) The present invention outputs the probability value, ranking result and confidence information of candidate anomaly causes, and links the results with recovery actions, log supplementation, work order generation and manual review process, thereby reducing the manual troubleshooting cost in large-scale fleet operation and maintenance and improving fault handling efficiency. In summary, this invention is applicable to vehicle networking environments with weak networks, frequent network switching, terminal restarts, or unstable log upload links, and has strong engineering practicality and large-scale deployment value. Attached Figure Description
[0025] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings: Figure 1 This is a flowchart of a method for compensating for abnormal vehicle network connections in a scenario with missing logs, according to the present invention. Figure 2 This is a block diagram of a vehicle network connection anomaly probability compensation positioning system for a scenario with missing logs, according to the present invention. Detailed Implementation
[0026] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0027] like Figure 1As shown, a method for compensating for and locating connection anomalies in vehicle-to-everything (V2X) scenarios with missing logs is deployed on a V2X cloud platform, connection management platform, or operation and maintenance analysis platform, including: Step S1: Obtain multi-source data related to connection anomalies of the target vehicle-to-everything (V2X) terminal within a preset time window; wherein, the multi-source data includes at least one or more of the following: terminal status data, network connection data, platform access data, and historical connection records.
[0028] Step S2: Perform integrity checks on the multi-source data to determine whether there are any missing logs related to connection anomalies, and determine the type and extent of missing logs; among which, the types of missing logs include, but are not limited to, missing attachment logs, missing authentication logs, missing heartbeat logs, missing network switching logs, missing eSIM status logs, and missing platform access logs.
[0029] Step S3: When it is determined that there are missing logs, extract the adjacent time-series status features of the target vehicle network terminal within a preset time period before and after the anomaly occurs; wherein, the adjacent time-series status features include, but are not limited to, the time of the last successful heartbeat before the anomaly, the most recent successful registration result, the number of reconnections before the anomaly, and the most recent network switching result.
[0030] Step S4: Extract the historical successful connection features of the target vehicle-to-everything (V2X) terminal; wherein, the historical successful connection features include, but are not limited to, the operator identifier, access point name, IP allocation result, connection establishment duration, continuous online duration, network type, and signal strength distribution in the most recent successful session.
[0031] Step S5: Filter the reference terminal group that has preset similar conditions to the target vehicle-to-everything (V2X) terminal and extract the group's statistical characteristics; among which, similar conditions include, but are not limited to, one or more of the following: same region, same operator, same vehicle model, same terminal version, same module version, and same time window.
[0032] In this embodiment, the reference terminal group can be selected from the terminal set managed by the vehicle network cloud platform, connection management platform, or operation and maintenance analysis platform. Specifically, it can be selected from historical terminal records and real-time online terminal records that are in the same or similar statistical range as the target vehicle network terminal. The statistical range includes one or more of the following: same region, same operator, same vehicle model, same terminal model, same terminal software version, same communication module version, same eSIM configuration, same time window, and same network standard.
[0033] Specifically, the vehicle-to-everything (V2X) cloud platform, connection management platform, or operation and maintenance analysis platform determines multiple reference terminals that meet similar conditions from the terminal database, connection session database, abnormal event database, or historical operation and maintenance sample database based on preset screening conditions. Then, it statistically analyzes the registration failure rate, authentication failure rate, disconnection rate, number of reconnections, signal strength distribution, or network switching anomaly ratio of the reference terminals in the corresponding time period as group statistical characteristics.
[0034] Step S6: Construct a set of compensation features based on the characteristics of adjacent time series states, historical successful connections, and group statistical features.
[0035] In another embodiment, in addition to the proximity time-series status, historical successful connection characteristics and group statistical characteristics described herein, the sources of compensation features may further include geofence information, weather information, base station handover information, vehicle ignition status, power supply status or server link quality indicators.
[0036] Step S7: Based on the compensation feature set, calculate the anomaly score of each cause in the candidate anomaly cause set.
[0037] In the embodiments, the candidate causes of anomalies include at least one or more of the following: terminal-side anomalies, communication module anomalies, SIM or eSIM anomalies, network access anomalies, operator-side anomalies, and platform-side access anomalies. Let the set of candidate anomaly causes be . Based on the compensation feature set, the matching degree between each preset candidate anomaly cause and the compensation feature set is evaluated to obtain the candidate anomaly causes. Corresponding abnormal scores ; Among them, the compensation features are derived from the features of adjacent time-series states, the features of historical successful connections, and the statistical features of the reference terminal group; Abnormal scores satisfy: ; In the formula, For candidate anomaly causes, use adjacent time-series state features The level of support Historical successful connection features for candidate anomaly causes The level of support To identify candidate anomaly causes based on the statistical characteristics of the terminal group The level of support Preset weighting coefficients for adjacent temporal state features. Preset weighting coefficients for historically successful connection features. To reference the preset weighting coefficients of the statistical characteristics of the terminal group, , and All are greater than 0.
[0038] Step S8: Normalize the anomaly score into a probability value to generate an anomaly localization result that includes candidate cause ranking, probability value, and confidence information.
[0039] In the embodiment, the anomaly score of each candidate anomaly cause is calculated. Normalization yields the probability value : .
[0040] Step S9: Trigger the handling process according to the probability value sorting; the handling process includes automatic recovery, log supplementation, re-issuance of configuration, generation of work order and transfer to manual review.
[0041] In the embodiment, when When the value exceeds a preset threshold, it can be directly triggered based on the candidate anomaly cause. Corresponding automatic recovery actions; when When the value is below the preset threshold, the process of supplementing logs, re-issuing configurations, generating work orders, or transferring the process to manual review is triggered.
[0042] like Figure 2 As shown, a vehicle network connection anomaly probability compensation and positioning system is deployed in a vehicle network cloud platform, connection management platform or operation and maintenance analysis platform, which implements a method for compensating and locating vehicle network connection anomalies in scenarios with missing logs. Specifically, it includes: anomaly detection module 1, data acquisition module 2, integrity judgment module 3, compensation feature construction module 4, probability positioning module 5, and result output and processing linkage module 6.
[0043] Anomaly detection module 1 is used to detect whether the target vehicle has connection anomalies. Anomalies include, but are not limited to, prolonged offline status, interrupted heartbeat, registration failure, authentication failure, frequent reconnection, or loss of connection after switching networks.
[0044] Data acquisition module 2 is used to acquire multi-source data related to connection anomalies within a preset time window when connection anomalies exist. The multi-source data includes at least one or more of the following: terminal status data, network connection data, platform access data, and historical connection records.
[0045] Integrity determination module 3 is used to identify the type and degree of missing logs and determine whether to enter the compensation and location process.
[0046] The compensation feature construction module 4 is used to construct a set of compensation features based on the adjacent time-series status, historical successful connection features, and statistical features of similar vehicle groups when there are missing logs. Among them, the historical successful connection features and statistical features of similar vehicle groups are all from the vehicle network terminals (including terminal database, connection session database, abnormal event database, and historical operation and maintenance sample library).
[0047] The probability localization module 5 is used to calculate the probability value of multiple candidate anomalies, generate anomaly localization results that include the candidate cause ranking, probability value and confidence information, and sort them according to the probability value.
[0048] The result output and processing linkage module 6 is used to trigger recovery actions, supplement logs, generate work orders, or transfer to manual review based on the sorting results.
[0049] Finally, it should be noted that the above are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.
Claims
1. A log loss scenario of a vehicle networking connection abnormal probability compensation positioning method, characterized in that, include: Step S1: Obtain multi-source data related to connection anomalies of the target vehicle-to-everything (V2X) terminal within a preset time window; Step S2: Perform integrity checks on the multi-source data to determine whether there are missing logs related to connection anomalies, and determine the type and extent of missing logs; Step S3: When it is determined that there are missing logs, extract the adjacent temporal state features of the target vehicle network terminal within a preset time period before and after the anomaly occurred; Step S4: Extract the historical successful connection features of the target vehicle-to-everything (V2X) terminal; Step S5: Filter the reference terminal group that has preset similar conditions to the target vehicle-to-everything (V2X) terminal and extract the group's statistical characteristics; Step S6: Construct a set of compensation features based on the characteristics of adjacent time series states, historical successful connections, and group statistical features; Step S7: Based on the compensation feature set, calculate the anomaly score of each cause in the candidate anomaly cause set; Step S8: Normalize the anomaly score into a probability value to generate an anomaly localization result that includes candidate cause ranking, probability value, and confidence information; Step S9: Trigger the handling process according to the probability value sorting; the handling process includes automatic recovery, log supplementation, re-issuance of configuration, generation of work order and transfer to manual review. 2.The log loss scenario of the Internet of Vehicles connection abnormal probability compensation positioning method according to claim 1, characterized in that, In step S1, the multi-source data includes at least one or more of the following: terminal status data, network connection data, platform access data, and historical connection records. 3.The log loss scenario of the Internet of Vehicles connection abnormal probability compensation positioning method according to claim 1, characterized in that, In step S2, the missing log types include, but are not limited to, missing attachment logs, missing authentication logs, missing heartbeat logs, missing network switching logs, missing eSIM status logs, and missing platform access logs. 4.The log loss scenario of the Internet of Vehicles connection anomaly probability compensation positioning method according to claim 1, characterized in that, In step S3, the adjacent time-series state features include, but are not limited to, the time of the last successful heartbeat before the anomaly, the most recent successful registration result, the number of reconnections before the anomaly, and the most recent network switching result. 5.The log loss scenario of a connected vehicle connection abnormal probability compensation positioning method according to claim 1, characterized in that, In step S4, the characteristics of historical successful connections include, but are not limited to, the operator identifier, access point name, IP allocation result, connection establishment duration, continuous online duration, network type, and signal strength distribution in the most recent successful session. 6.The log loss scenario of a connected vehicle connection abnormality probability compensation positioning method according to claim 1, characterized in that, In step S5, similar conditions include, but are not limited to, one or more of the following: same region, same operator, same vehicle model, same terminal version, same module version, and same time window.
7. The log loss scenario of the Internet of Vehicles connection exception probability compensation positioning method according to claim 1, characterized in that, In step S7, the candidate abnormality reasons include at least one or more of the following: terminal-side abnormality, communication module abnormality, SIM or eSIM abnormality, network access abnormality, operator-side abnormality, and platform-side access abnormality. The candidate abnormal reason set is , based on the compensation feature set, respectively evaluating the matching degree of each preset candidate abnormal reason and the compensation feature set, obtaining each candidate abnormal reason corresponding to the abnormal score ; Among them, the compensation features are derived from the features of adjacent time-series states, the features of historical successful connections, and the statistical features of the reference terminal group; Abnormal scores satisfy: ； In the formula, a support degree of the candidate abnormal cause by the adjacent time sequence state feature, a support degree of the candidate abnormal cause by the historical successful connection feature, a support degree of the candidate abnormal cause by the reference terminal group statistical feature, a preset weight coefficient of the adjacent time sequence state feature, a preset weight coefficient of the historical successful connection feature, a preset weight coefficient of the reference terminal group statistical feature, and are all greater than 0. 8. The method for compensating for abnormal vehicle network connection probability in a log missing scenario according to claim 7, characterized in that, In step S8, the anomaly score of each candidate anomaly cause is calculated. Normalization yields the probability value : 。 9. The method for compensating for abnormal vehicle network connections in a log-missing scenario according to claim 8, characterized in that, In step S9 when When the value exceeds a preset threshold, it can be directly triggered based on the candidate anomaly cause. The corresponding automatic recovery action; when When the value is below the preset threshold, the process of supplementing logs, re-issuing configurations, generating work orders, or transferring the process to manual review is triggered.
10. A vehicle-to-everything (V2X) connection anomaly probability compensation positioning system implementing the method of claims 1-9, characterized in that, include: The anomaly detection module is used to detect whether the target vehicle has connection anomalies. Anomalies include, but are not limited to, prolonged offline status, interrupted heartbeat, registration failure, authentication failure, frequent reconnection, or loss of connection after switching networks. The data acquisition module is used to acquire multi-source data related to connection anomalies within a preset time window when connection anomalies occur. The multi-source data includes at least one or more of the following: terminal status data, network connection data, platform access data, and historical connection records. The integrity determination module is used to identify the type and extent of missing logs and determine whether to proceed with the compensation and location process. The compensation feature construction module is used to construct a set of compensation features based on the adjacent time-series status, historical successful connection features, and statistical features of similar vehicle groups when there are missing logs. The probability localization module is used to calculate the probability value of multiple candidate anomalies, generate anomaly localization results that include the candidate cause ranking, probability value and confidence information, and sort them according to the probability value. The results output and processing linkage module is used to trigger recovery actions, supplement logs, generate work orders, or transfer to manual review based on the sorting results.